Deep Learning-Based Brain Tumor Segmentation—An Overview

Brain tumor segmentation is a subdivision of study that focuses on creating a precise delineation of brain tumor regions using magnetic resonance imaging (MRI). Nowadays, deep learning (DL) algorithms have shown remarkable growth in addressing a range of computer vision problems like semantic segmentation, image classification, and object detection. Applying DL techniques to electronic health records can yield useful insights that can be used to improve patient risk score systems or identify the onset of a brain tumor. Deep learning techniques can also facilitate the effective processing and balanced assessment of the vast volumes of MRI-based image data. This paper presents an overview of the traditional and deep learning techniques used for brain tumor segmentation (BTS) and brain tumor classification (BTC). Further, a detailed analysis of the existing surveys was carried out to determine the overall impact of each survey. In addition, this study discusses the deep learning-based methods with their respective classification/segmentation, features, pre-processing, and outcomes. Finally, this study not only examines the prior research but identifies potential future research directions.